Humanoid robots generated a revenue of only 823,000 yuan last year. Is it reasonable that DeepRobotics' price-to-sales ratio is even higher than that of Unitree?

2026 is regarded as the first year of mass production of humanoid robots, with the expectation of transitioning from "performance toys" to working in factories. Tesla's Optimus Gen3 is confirmed to enter mass production by the end of 2026, with a planned annual production capacity of 1 million units. Domestic companies such as Unitree, Zhipu, and Ubtech are making bulk shipments.

As one of the "Six Rising Stars in Hangzhou," Deep Robotics was established on November 29, 2017, with its headquarters in Xihu District, Hangzhou. The company focuses on the innovation and industrial application of embodied intelligence technology, and is dedicated to the R & D, production, sales, and services of quadruped robots, humanoid robots, and core components.

Not long ago, the prospectus submitted by Deep Robotics showed that the company successfully turned a profit in 2025. Its revenue mainly came from quadruped robots and wheel - legged robots. Only one unit of the DR series humanoid robots was sold last year, with an income of 823,000 yuan, accounting for only 0.24% of the total revenue.

In contrast, Unitree shipped over 5,500 humanoid robots in 2025, with an income of 877 million yuan.

In terms of performance, Deep Robotics achieved a revenue of 3.37 billion yuan last year. According to the financing amount and the corresponding share ratio disclosed in its prospectus, the corresponding market value of Deep Robotics after listing is 13.9 billion yuan, corresponding to a price - to - sales ratio of about 41 times. In comparison, Unitree Technology, which was also sprinting for an IPO during the same period, had a revenue of 1.708 billion yuan and a valuation of 42 billion yuan, with a price - to - sales ratio of about 25 times. This means that Deep Robotics is about 60% more expensive in terms of valuation multiples than Unitree.

Many market insiders believe that humanoid robots are more likely to be the ultimate form, carrying a more grand capital narrative. Deep Robotics lags far behind leading companies in this field such as Unitree.

The founder, Zhu Qiuguo, is a doctoral supervisor at Zhejiang University

As early as 2025, Deep Robotics began to plan for an IPO.

The public issuance counseling publicity system of the China Securities Regulatory Commission shows that on December 23 last year, Deep Robotics completed the IPO counseling filing, with CITIC Construction serving as the counseling institution. It will submit an application for counseling acceptance from April to June 2026, cooperate to complete the counseling acceptance, and at the same time prepare for the production of application documents for the initial public offering of stocks.

At the beginning of May this year, Deep Robotics completed the IPO counseling.

On May 18, Deep Robotics officially disclosed its prospectus. The company plans to issue no less than 82.98 million shares, accounting for no less than 18% of the total share capital after issuance, and raise about 2.5 billion yuan. The funds will be invested in projects such as the R & D of embodied algorithms and models, the R & D of robot bodies and solutions, the industrialization of embodied intelligent robots, and the construction of an embodied intelligent robot base.

Leida Finance noticed that the key figures of the company are Zhu Qiuguo and Li Chao.

Public information shows that in November 2017, Zhu Qiuguo and Li Chao jointly founded Deep Robotics. Both of them are from the laboratory of Zhejiang University.

As of the date of prospectus disclosure, Zhu Qiuguo serves as the chairman and general manager of Deep Robotics, and Li Chao serves as a director and deputy general manager of the company.

Zhu Qiuguo directly holds 1.35 million shares of Deep Robotics, with a shareholding ratio of 15.61%. He indirectly controls 10.23% of the company's equity through the employee shareholding platform Hangzhou Kongjian Investment Management Partnership (Limited Partnership). Li Chao holds 5.46% of the company's equity. He has signed a concerted action agreement with Zhu Qiuguo and is Zhu Qiuguo's concerted action person, keeping consistent with Zhu Qiuguo in the decision - making and actions of all major matters of the company.

Therefore, Zhu Qiuguo is the controlling shareholder and actual controller of the issuer, and he and his concerted action persons jointly control 31.05% of the issuer's shares.

Zhu Qiuguo is also a professor and doctoral supervisor at the School of Control Science and Engineering of Zhejiang University. The official website of Zhejiang University shows that Zhu Qiuguo was born in 1982. He is engaged in research on legged robot technology, intelligent robot learning and control, etc. He has successively undertaken projects such as the national key R & D program, the National Natural Science Foundation, and the key R & D program of Zhejiang Province. He won the first - class award of the Science and Technology Award of Zhejiang Province in 2013, the first - class award of the Teaching Achievement Award of Zhejiang Province in 2014, the second - class award of the National Teaching Achievement Award in 2018, the first - class award of the Science and Technology Progress Award of Zhejiang Province in 2024, and the first - class award of the Teaching Achievement Award of Zhejiang Province in 2025.

Li Chao graduated from Tianjin Polytechnic University and then obtained a doctorate from Zhejiang University. He is also an off - campus cooperative tutor for master's students at Zhejiang University and has more than 15 years of experience in robot technology and product R & D. His main research field is the planning and control of humanoid robots.

It is reported that Zhu Qiuguo's robot journey began when he participated in the RoboCup robot world cup competition during his undergraduate years. These experiences gave him the opportunity to transform professional theoretical knowledge into practice.

In addition, in 2016, Boston Dynamics in the United States demonstrated the quadruped robot SpotMini, which weighed only 25 kilograms. The wonderful demonstration video posted on the Internet refreshed the world's inherent impression of robots. Zhu Qiuguo later recalled that in those years, Boston Dynamics frequently released various robot stunt videos: "Each one seemed to be reminding us that China didn't have such robots yet."

However, after officially embarking on entrepreneurship, even he, with a professional background, quickly realized that it was still very difficult to develop a robot product with stable performance. In an interview in September 2025, Zhu Qiuguo revealed that in the early days of entrepreneurship, he and his team tried countless versions of the robot operating system. "The motors were wound by hand, and we had to develop them ourselves to improve the power density."

Fortunately, after a period of exploration, Deep Robotics successfully launched the first - generation quadruped robot "Jueying X10," which was the first domestic quadruped robot capable of going up and down stairs and autonomous navigation at that time.

In 2018, with its first - generation robot dog product, Deep Robotics received millions of yuan in external investment. Tianyancha shows that after the angel - round financing in July 2018, Deep Robotics didn't conduct the Series A financing until 2020. So far, the company has completed 8 rounds of financing. On December 25 last year, the company announced the completion of hundreds of millions of yuan in Pre - IPO round financing.

This round of financing was led by the National Artificial Intelligence Industry Fund, with a large Internet company following. This was also the third round of financing officially announced by Deep Robotics in the second half of 2025.

Only one humanoid robot was sold last year

In recent years, there has been a boom in hard - tech entrepreneurship represented by fields such as GPU chips, large AI models, embodied intelligence, and commercial space. Most of these projects have similar characteristics, that is, large R & D investment, long time cycles, and widespread losses in the early stage of development. They especially rely on blood transfusions from the primary and secondary markets to gradually form technological barriers.

Therefore, hard - tech projects not only compete in technological R & D capabilities but also in financing capabilities. However, as the industry moves from the "story - telling" stage to the capital - market listing, competing in financing capabilities is only the starting point. Enterprises have officially entered a new stage centered on technological strength, revenue - generating capabilities, and sustainable business models.

For Deep Robotics, it has relatively strong capabilities in the field of embodied intelligence. In September 2025, Zhu Qiuguo told the media that the company's revenue in 2024 increased by more than 100% compared with that in 2023. The company's robot shipments will reach the level of 10,000 units in 2025, and it has now entered markets in Asia - Pacific, the Middle East, Europe, and the United States.

In terms of products, Deep Robotics claims that the Lynx M20, released in April 2025, is the world's first industry - application wheel - legged robot designed for complex terrains and dangerous environments. It weighs 33 kg, has an IP66 protection rating, can operate normally in a temperature range of - 20°C to 55°C, and can climb an 80 - centimeter high platform.

In October of the same year, the humanoid robot DR02 released by Deep Robotics has an IP66 protection rating and can operate stably in a wide - temperature environment of - 20°C to 55°C, breaking the limitation that humanoid robots can only be used in controlled environments. Its walking speed reaches 1.5 m/s, and the maximum speed is 4 m/s. It can easily cross a 25 - cm high staircase.

With the improvement of product performance, the embodied intelligence of Deep Robotics has gradually entered the stage of commercial application. In the prospectus, Deep Robotics revealed that the company continues to deepen its full - scenario layout. Its industry applications have covered core fields such as power inspection, emergency fire protection, industrial inspection, police security, public infrastructure inspection, patrol, and logistics transportation. The speed, breadth, and depth of scenario expansion continue to increase. Currently, the number of the company's industry - level application customers exceeds 500, and its products are exported to more than 45 countries or regions around the world, including the United States, Singapore, and Germany. It has taken the lead in achieving large - scale implementation breakthroughs in multiple industry - application fields globally.

Currently, the main sources of Deep Robotics' revenue are quadruped robots and wheel - legged robots.

The prospectus shows that in 2025, Deep Robotics ranked first in the world in terms of revenue from the industry - application field of quadruped robots and second in the world in terms of quadruped robot revenue. The combined revenue of the two last year reached 322 million yuan, accounting for more than 95% of the total revenue.

In 2025, Deep Robotics sold only one humanoid robot, less than the three units sold in 2024. The income was only 823,000 yuan, accounting for only 0.24% of the total revenue.

In contrast, Unitree shipped over 5,500 humanoid robots in 2025, with an income of 877 million yuan; Zhipu shipped 5,168 units; and Ubtech's production capacity target in 2026 is in the tens of thousands.

The revenue structure of Deep Robotics may be closely related to Zhu Qiuguo's perception.

Regarding quadruped robots (robot dogs), Zhu Qiuguo said in a dialogue with the media in August last year that in recent years, AI technology, especially technologies such as reinforcement learning, has continuously enhanced the technical capabilities of robot dogs and their adaptability to scenarios. In this sense, the "technological gap" in the robot - dog track has generally been crossed. However, this doesn't mean that every company has crossed it. Each company needs to find its own application scenarios, and the products need to be truly stable and reliable and able to bring real value to customers.

Regarding humanoid robots, Zhu Qiuguo said at that time that he didn't deny their long - term value. They may play a major role in 10 or 20 years, but many current discussions are just gimmicks. It is still in the early stage of technological exploration and far from mature for commercialization.

In terms of performance, Deep Robotics had a revenue of 337 million yuan and a net profit attributable to the parent company of 28.684 million yuan in 2025, achieving annual profitability for the first time. The net profit attributable to the parent company after deducting non - recurring gains and losses was 15.1232 million yuan.

Based on a revenue of 337 million yuan and a valuation of 1.39 billion yuan, Deep Robotics has a price - to - sales ratio of about 41 times. In comparison, Unitree Technology, which was also sprinting for an IPO during the same period, had a revenue of 1.708 billion yuan and a valuation of 4.2 billion yuan, with a price - to - sales ratio of about 25 times. This means that Deep Robotics is about 60% more expensive in terms of valuation multiples than Unitree.

Some market insiders believe that humanoid robots are more likely to be the ultimate form, and the industry has entered the stage of large - scale production. Deep Robotics lags far behind the leading companies in the humanoid - robot field. In this case, if Deep Robotics doesn't increase its investment, it will be difficult to seize the dividends of the general humanoid - robot main track.

This article is from the WeChat official account "Leida Finance", written by Li Yihui and edited by Shen Hai. It is reprinted by 36Kr with permission.